mlsanityData to 7 Oct 2026

Sim-to-Real of Humanoid Locomotion Policies via Joint Torque Space Perturbation Injection

arXiv 2603.21853 ↗First posted 23 Mar 2026cs.RO cs.AIcited by 0

Not on the map. Why?

Abstract (arXiv)

This paper proposes a novel alternative to existing sim-to-real methods for training control policies with simulated experiences. Unlike prior methods that typically rely on domain randomization over a fixed finite set of parameters, the proposed approach injects state-dependent perturbations into the input joint torque during forward simulation. These perturbations are designed to simulate a broader spectrum of reality gaps than standard parameter randomization without requiring additional training. By using neural networks as flexible perturbation generators, the proposed method can represent complex, state-dependent uncertainties, such as nonlinear actuator dynamics and contact compliance, that parametric randomization cannot capture. Experimental results demonstrate that the proposed approach enables humanoid locomotion policies to achieve superior robustness against complex, unseen reality gaps in both simulation and real-world deployment.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 31 of its 43 references that are in the map. First posted on arXiv 23 Mar 2026.

Builds on

31 of 43 references.

References · 31

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,401
  3. 2018Soft Actor-Critic Algorithms and Applicationscited by 1,179
  4. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 882
  5. 2017Sim-to-Real Transfer of Robotic Control with Dynamics Randomizationcited by 616
  6. 2021RMA: Rapid Motor Adaptation for Legged Robotscited by 430
  7. 2019Learning agile and dynamic motor skills for legged robotscited by 402
  8. 2021Learning to Walk in Minutes Using Massively Parallel Deep Reinforcement Learningcited by 395
  9. 2022DayDreamer: World Models for Physical Robot Learningcited by 372
  10. 2018Sim-to-Real: Learning Agile Locomotion For Quadruped Robotscited by 277
  11. 2020Learning Agile Robotic Locomotion Skills by Imitating Animalscited by 229
  12. 2021How to Train Your Robot with Deep Reinforcement Learning; Lessons We've Learnedcited by 201
19 more
  1. 2022A Walk in the Park: Learning to Walk in 20 Minutes With Model-Free Reinforcement Learningcited by 76
  2. 2020Sim-to-Real Learning of All Common Bipedal Gaits via Periodic Reward Compositioncited by 71
  3. 2022Adversarial Motion Priors Make Good Substitutes for Complex Reward Functionscited by 68
  4. 2021Minimizing Energy Consumption Leads to the Emergence of Gaits in Legged Robotscited by 67
  5. 2024Reinforcement Learning for Versatile, Dynamic, and Robust Bipedal Locomotion Controlcited by 63
  6. 2018Learning Symmetric and Low-energy Locomotioncited by 53
  7. 2024Advancing Humanoid Locomotion: Mastering Challenging Terrains with Denoising World Model Learningcited by 38
  8. 2022GenLoco: Generalized Locomotion Controllers for Quadrupedal Robotscited by 36
  9. 2024Learning Smooth Humanoid Locomotion through Lipschitz-Constrained Policiescited by 26
  10. 2023HumanMimic: Learning Natural Locomotion and Transitions for Humanoid Robot via Wasserstein Adversarial Imitationcited by 23
  11. 2020Crossing The Gap: A Deep Dive into Zero-Shot Sim-to-Real Transfer for Dynamicscited by 20
  12. 2024Learning Humanoid Locomotion over Challenging Terraincited by 20
  13. 2020Policy Transfer via Kinematic Domain Randomization and Adaptationcited by 16
  14. 2020Learning Task Space Actions for Bipedal Locomotioncited by 15
  15. 2022Learning and Deploying Robust Locomotion Policies with Minimal Dynamics Randomizationcited by 15
  16. 2023Torque-based Deep Reinforcement Learning for Task-and-Robot Agnostic Learning on Bipedal Robots Using Sim-to-Real Transfercited by 12
  17. 2019Modelling Generalized Forces with Reinforcement Learning for Sim-to-Real Transfercited by 5
  18. 2024Extended Friction Models for the Physics Simulation of Servo Actuatorscited by 3
  19. 2023A Model Predictive Capture Point Control Framework for Robust Humanoid Balancing via Ankle, Hip, and Stepping Strategiescited by 1

Built on by

Not cited yet.